摘要目的:建立一种预测创伤性下肢骨折后深静脉血栓形成（DVT）风险的指标模型。方法:对2019年5月至12月北京积水潭医院的424例创伤性下肢骨折患者进行巢式病例对照研究。创伤性下肢骨折患者的年龄≥18岁；下肢骨折包括髋部骨折、股骨干骨折、股骨远端骨折、髌骨骨折、胫骨平台骨折、胫腓骨骨干骨折、踝关节骨折和足部骨折。创伤患者在术前行静脉造影检查，424例创伤性下肢骨折患者中，56例在术前诊断为DVT；在非DVT患者中挑选与DVT患者年龄、性别、骨折部位相匹配的对照。检测指标包括常规凝血试验和血栓分子标志物，如D-二聚体、纤溶酶-ɑ2-抗纤溶酶复合物（PIC）、组织型纤溶酶原激活剂-纤溶酶原激活剂抑制物复合物（tPAIC）等，并计算Caprini评分。将DVT患者和非DVT患者的指标进行单因素分析，再利用logistic回归分析，筛选创伤后DVT发生的独立危险因素。应用受试者工作特征（ROC）曲线分析，评估各指标预测创伤后DVT风险的效能。结果:创伤性下肢骨折后发生DVT患者的tPAIC水平和Caprini评分明显大于非DVT患者（ P值分别为0.036和0.016）。D-二聚体、PIC、Caprini评分是创伤性下肢骨折后术前DVT发生的独立危险因素。利用上述3个独立危险因素计算新指标：预测因子 = 0.098×D-二聚体（mg/l FEU）+（-0.564）×PIC（μg/ml）+0.233×Caprini评分。预测因子的ROC曲线下面积最大为0.721。 结论:预测因子，由D-二聚体、PIC和Caprini评分计算而得，可较综合地涵盖创伤性下肢骨折后发生DVT患者和未发生DVT患者之间各类危险因素的变化，能较好地预测创伤性下肢骨折后DVT风险。

abstractsObjective:To seek indicators or models for predicting the risk of deep vein thrombosis (DVT) after traumatic lower extremity fractures.Methods:A nested case-control study was conducted on 424 patients with traumatic lower extremity fractures in Beijing Jishuitan Hospital from May 2019 to December 2019. Patients with traumatic lower extremity fractures of the hip, femoral shaft, distal femoral, patella, tibial plateau, tibiofibular shaft, ankle or foot were 18 or older. Trauma patients were examined by venogram before surgery. 56 of 424 patients with traumatic lower limb fractures were diagnosed with preoperative DVT. Then, those with no evidence of DVT were paired by age, gender and fracture sites to those with DVT. Laboratory parameters included conventional coagulation tests and thrombotic biomarkers, such as D-dimer, plasmin-ɑ2-plasmin inhibitor complex (PIC), tissue plasminogen activator-plasminogen activator inhibitor complex (tPAIC), and Caprini score was calculated. Univariate analysis was used to distinguish statistically significant variables between trauma patients with and without DVT. Then logistic regression analysis was applied to identify independent risk factors for post-traumatic DVT. The receiver-operating characteristic (ROC) curve analysis was employed to assess the efficacy of indicators in predicting the risk of DVT after traumatic lower extremity fractures.Results:The levels of tPAIC and Caprini score in patients with DVT after traumatic lower limb fractures were significantly higher than those in non-DVT patients ( P values of 0.036 and 0.016, respectively). D-dimer, PIC, and Caprini score were independent risk factors for preoperative DVT after traumatic lower limb fractures. Thus, we defined the Predicting Index as 0.098×D-dimer (mg/L FEU)+(-0.564) ×PIC (μg/ml)+0.233×Caprini score, and found that the area under the ROC curve for the Predicting Index was 0.721. Conclusions:Predicting Index, calculated by D-dimer, PIC and Caprini score, can comprehensively encompass the changes of risk factors between patients with and without DVT after traumatic lower limb fractures, and can well predict the risk of DVT after traumatic lower extremity fractures.